A phenomenological Vacuum Mass Fraction (VMF) model is proposed, in which the hadron mass is separated into two components: (1) the current mass, determined by the explicit breaking of chiral symmetry via sigma terms, and (2) the residual nonperturbative component M*, generated by confinement, gluon field energy, and the QCD trace anomaly. Using lattice QCD data for the pion-nucleon (σN ≈ 44 MeV) and strange (σsN ≈ 30 MeV) sigma terms, the vacuum value of the nonperturbative component is fixed at M*,0 = 859 ± 8 MeV, constituting ~91% of the nucleon mass. A series of EOS prototypes for dense nuclear matter is constructed with a two-parameter ansatz for the in-medium modification M*(nB). It is shown that (i) minimal realizations with constant vector repulsion yield a superluminal speed of sound; (ii) density-dependent saturation of the vector field resolves the causality problem (cs² < 1); (iii) the problem of an overly stiff equation of state is resolved by introducing a first-order phase transition to the conformal QGP limit, reducing the maximum neutron star mass to a realistic Mmax ≈ 2.3 M⊙. The model is strictly bounded by input data: the vacuum scale M*,0 is fixed by lattice calculations and is not a free fitting parameter.
Oleg Kirichenko (2026) studied this question.